Trait TensorIndexing 

pub trait TensorIndexing: TensorStructural {
    // Required methods
    fn gather(
        &mut self,
        operand: &Tensor,
        start_indices: &Tensor,
        config: &GatherConfig,
    ) -> Result<Tensor>;
    fn scatter(
        &mut self,
        operand: &Tensor,
        scatter_indices: &Tensor,
        updates: &Tensor,
        config: &ScatterConfig,
    ) -> Result<Tensor>;
    fn slice(&mut self, input: &Tensor, config: &SliceConfig) -> Result<Tensor>;
    fn dynamic_slice(
        &mut self,
        input: &Tensor,
        starts: &Tensor,
        slice_sizes: &[usize],
    ) -> Result<Tensor>;
    fn dynamic_update_slice(
        &mut self,
        operand: &Tensor,
        update: &Tensor,
        starts: &Tensor,
    ) -> Result<Tensor>;
    fn pad(&mut self, input: &Tensor, config: &PadConfig) -> Result<Tensor>;
    fn concatenate(&mut self, inputs: &[&Tensor], axis: usize) -> Result<Tensor>;
    fn reverse(&mut self, input: &Tensor, axes: &[usize]) -> Result<Tensor>;

    // Provided methods
    fn scatter_read(
        &mut self,
        operand: TensorRead<'_>,
        scatter_indices: TensorRead<'_>,
        updates: TensorRead<'_>,
        config: &ScatterConfig,
    ) -> Result<Tensor> { ... }
    fn slice_read(
        &mut self,
        input: TensorRead<'_>,
        config: &SliceConfig,
    ) -> Result<Tensor> { ... }
    fn dynamic_slice_read(
        &mut self,
        input: TensorRead<'_>,
        starts: TensorRead<'_>,
        slice_sizes: &[usize],
    ) -> Result<Tensor> { ... }
    fn dynamic_update_slice_read(
        &mut self,
        operand: TensorRead<'_>,
        update: TensorRead<'_>,
        starts: TensorRead<'_>,
    ) -> Result<Tensor> { ... }
    fn pad_read(
        &mut self,
        input: TensorRead<'_>,
        config: &PadConfig,
    ) -> Result<Tensor> { ... }
    fn concatenate_read(
        &mut self,
        inputs: &[TensorRead<'_>],
        axis: usize,
    ) -> Result<Tensor> { ... }
    fn reverse_read(
        &mut self,
        input: TensorRead<'_>,
        axes: &[usize],
    ) -> Result<Tensor> { ... }
}

Indexing, slicing, and padding operations.

Examples

use tenferro_tensor::TensorIndexing;

fn accepts_indexing<B: TensorIndexing>(_backend: &mut B) {}

Required Methods

fn gather( &mut self, operand: &Tensor, start_indices: &Tensor, config: &GatherConfig, ) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn scatter( &mut self, operand: &Tensor, scatter_indices: &Tensor, updates: &Tensor, config: &ScatterConfig, ) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn slice(&mut self, input: &Tensor, config: &SliceConfig) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. In particular, a limit greater than the corresponding input dimension is reported as crate::ValidationError::InvalidArgument with the "configuration" argument. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn dynamic_slice( &mut self, input: &Tensor, starts: &Tensor, slice_sizes: &[usize], ) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn dynamic_update_slice( &mut self, operand: &Tensor, update: &Tensor, starts: &Tensor, ) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn pad(&mut self, input: &Tensor, config: &PadConfig) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn concatenate(&mut self, inputs: &[&Tensor], axis: usize) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn reverse(&mut self, input: &Tensor, axes: &[usize]) -> Result<Tensor>

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

Provided Methods

fn scatter_read( &mut self, operand: TensorRead<'_>, scatter_indices: TensorRead<'_>, updates: TensorRead<'_>, config: &ScatterConfig, ) -> Result<Tensor>

Scatter with every input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, ScatterConfig, TensorRead};

fn run(session: &mut dyn TensorIndexing, x: TensorRead<'_>, idx: TensorRead<'_>, updates: TensorRead<'_>, config: &ScatterConfig)
    -> tenferro_tensor::Result<Tensor>
{
    session.scatter_read(x, idx, updates, config)
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn slice_read( &mut self, input: TensorRead<'_>, config: &SliceConfig, ) -> Result<Tensor>

Slice with the input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, SliceConfig, TensorRead};

fn run(session: &mut dyn TensorIndexing, x: TensorRead<'_>, config: &SliceConfig)
    -> tenferro_tensor::Result<Tensor>
{
    session.slice_read(x, config)
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn dynamic_slice_read( &mut self, input: TensorRead<'_>, starts: TensorRead<'_>, slice_sizes: &[usize], ) -> Result<Tensor>

Dynamic slice with every input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, TensorRead};

fn run(session: &mut dyn TensorIndexing, x: TensorRead<'_>, starts: TensorRead<'_>)
    -> tenferro_tensor::Result<Tensor>
{
    session.dynamic_slice_read(x, starts, &[2])
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn dynamic_update_slice_read( &mut self, operand: TensorRead<'_>, update: TensorRead<'_>, starts: TensorRead<'_>, ) -> Result<Tensor>

Dynamic-update-slice with every input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, TensorRead};

fn run(
    session: &mut dyn TensorIndexing,
    x: TensorRead<'_>,
    update: TensorRead<'_>,
    starts: TensorRead<'_>,
) -> tenferro_tensor::Result<Tensor> {
    session.dynamic_update_slice_read(x, update, starts)
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn pad_read( &mut self, input: TensorRead<'_>, config: &PadConfig, ) -> Result<Tensor>

Pad with the input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, PadConfig, TensorRead};

fn run(session: &mut dyn TensorIndexing, x: TensorRead<'_>, config: &PadConfig)
    -> tenferro_tensor::Result<Tensor>
{
    session.pad_read(x, config)
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn concatenate_read( &mut self, inputs: &[TensorRead<'_>], axis: usize, ) -> Result<Tensor>

Concatenate with every input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, TensorRead};

fn run(session: &mut dyn TensorIndexing, a: TensorRead<'_>, b: TensorRead<'_>)
    -> tenferro_tensor::Result<Tensor>
{
    session.concatenate_read(&[a, b], 0)
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

fn reverse_read( &mut self, input: TensorRead<'_>, axes: &[usize], ) -> Result<Tensor>

Reverse with the input taken as a read.

The default stages each borrowed view into compact storage (an owned tensor is passed through) and delegates to the owned-input method; a backend whose kernel consumes a borrowed operand should override this so a view is not copied (#2214).

Examples

use tenferro_tensor::{Tensor, TensorIndexing, TensorRead};

fn run(session: &mut dyn TensorIndexing, x: TensorRead<'_>)
    -> tenferro_tensor::Result<Tensor>
{
    session.reverse_read(x, &[0])
}

Errors

Returns crate::Error::Validation with a typed ValidationError source for invalid shapes, ranks, axes, dtypes, or output metadata. It returns crate::Error::BackendFailure or crate::Error::BackendSource when backend execution or storage access cannot provide the requested result.

Dyn Compatibility

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors